Gemini Omni 1.1 Flash

Gemini Omni 1.1 Flash Text-to-Video API on E2X

Omni 1.1 Flash went generally available at Google on 27 August 2026 — a Gemini-family model, not a Veo. We run its prompt-only capability as google/omni-1.1-flash/text-to-video, billed per second of output: $0.09 at 720p, off a $0.15 list rate after a standing 40% discount. The default eight seconds is $0.72. A prompt is the whole payload.

One number here went up. The predecessor, Gemini Omni Flash text-to-video, bills $0.0825 a second on this platform; this page bills $0.09. Nine percent more, and you should read it here rather than infer it from an invoice. The nine percent does not buy a better second of video. It buys a $0.0297 draft tier the older model has no resolution enum for, clips that reach ten seconds, and two capabilities the predecessor never shipped on E2X at all. Our notes on the 1.1 release walk through the rest.

Now the sentence that argues against our own invoice. Veo 3.1 Fast text-to-video runs on this platform at $0.01 a second — a ninth of this page's price. One clip, from a sentence, nothing downstream depending on it? Go there. The premium here buys ten seconds, 4K, a cheap draft mode, and shots that belong to each other. Buy none of those and you have bought nothing.

Other material, other endpoint: a still becoming frame one goes to image-to-video, two ends with a gap between to start-end-frame-to-video, a repeating face to reference-to-video.

What a second of this model costs across providers

Every figure below was read on August 28, 2026 from the provider's own version 1.1 page:

Google's row is its own arithmetic, published on its own pricing page: $17.50 per million output tokens, 5,792 tokens in a second of 720p, $0.1014 — which is why the footnote there settles on "an effective price of approximately $0.10 per second." Every row includes sound, so the comparison is like for like. This model has no audio toggle at all: dialogue and room tone are generated with the picture, and nobody on that list bills them separately. Google publishes the same rate at every resolution, as a token count — 1,931 per second of 360p, rising to 5,792, 8,688 and 17,376 as you climb the tiers, with audio inside every one of them. So the ladder below compares like with like the whole way up.

Spot-check us, but read the version string. fal.ai keeps two listings: a versionless google/gemini-omni-flash page pricing the older preview weights near $0.125 a second, and a v1.1 page at $0.10. Ours is v1.1. Wavespeed earns the reverse caution — $0.13 is confirmed for its text-to-video variant alone, so its other figures stay out.

The 720p rate is not the interesting half. This is:

Our multiplier from 720p to 4K is 2×. Google's is exactly 3× — its own token counts go 5,792 to 17,376 — and the resellers track Google. At the bottom the two ladders are the same shape — 0.33× down to 360p on both sides — and they separate as you climb: by 4K Google has tripled its 720p rate while we have doubled ours. That is why no tier on this ladder has Google below us: we are twelve percent under at 360p, eleven under at 720p, the same eleven at 1080p, and forty-one percent under at 4K. Runware repays a glance too: level with Google at 720p, then it charges $0.16 for 1080p and $0.32 for 4K, above Google's own rate on both tiers. One note if you spot-check us: Google's announcement table rounds these to $0.03, $0.10, $0.15 and $0.30. The figures above come from its own per-resolution token counts, which is what the invoice is calculated on.

Prices and product terms can change. Check each provider's current pricing before making a purchasing decision. Google Batch or Flex pricing is not directly equivalent to a standard on-demand API request because scheduling, availability, and processing conditions differ; it is therefore excluded from this comparison. This is a scoped comparison, not a claim that E2X is the world's cheapest option in every configuration.

Draft at 360p, render the keeper at 720p

Repetition empties the account, not any one generation: attempt seven costs what attempt one did.

Four seconds at 360p is $0.1188; the same four at 720p is $0.36. Six drafts cheap plus one render at the default comes to $1.07, against $2.52 for seven full-price passes. Google claims 360p returns up to 60% faster too — its number, not ours.

Ten seconds of 4K, at $0.18 a second, is $1.80 — the largest single bill here.

Sending the request

Generate a key in the dashboard, keep it off the client, and post the job. prompt is the only required field.

curl -X POST https://api.e2x.ai/v1/jobs/submit \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/omni-1.1-flash/text-to-video",
    "input": {
      "prompt": "A welder lowers her mask, sparks throwing blue light across stacked steel. Slow dolly left. Arc crackle, a fan off frame.",
      "resolution": "360p",
      "duration": "6",
      "seed": 20260828
    }
  }'

That is a draft pass: six seconds at 360p, $0.1782. Switch resolution to 720p and it costs $0.54.

const headers = {
  Authorization: `Bearer ${process.env.E2X_API_KEY}`,
  "Content-Type": "application/json",
};

const { data } = await fetch("https://api.e2x.ai/v1/jobs/submit", {
  method: "POST",
  headers,
  body: JSON.stringify({
    model: "google/omni-1.1-flash/text-to-video",
    input: { prompt: "Rain hammers a bus shelter at night.", duration: "4" },
  }),
}).then((r) => r.json());

for (let attempt = 0; attempt < 360; attempt++) {
  const job = await fetch(`https://api.e2x.ai/v1/jobs/${data.jobId}`, { headers })
    .then((r) => r.json());
  const { status } = job.data;

  if (status === "completed") { await archive(job.data.outputs[0].url); break; }
  if (status === "failed" || status === "cancelled") throw new Error(status);
  await new Promise((r) => setTimeout(r, 5000));
}

Budget four minutes a job, or pass a webhookUrl. Two traps: duration and resolution travel as strings, "10" not 10; and the returned URL expires, so copy the file somewhere permanent as you read it. Every field sits in the machine-readable spec.

Which of our video models you should actually call

Cheapest first:

Read row one against row two. Nine times the price is a lot for a clip nobody will cut against anything; continuity is what the premium buys. Everything else that turns a prompt into footage is under text-to-video; every model we run is on one page.

Before you build on it

Output is fixed: 24 fps, three to ten seconds, 16:9 or 9:16. That ten matters, because Veo stops at eight. The model card names three failure modes:

"Maintaining complete consistency throughout edits, generating scenes with complex motion, or rendering perfectly accurate text remains a challenge."

On-screen text catches people out: ask for a shop sign and expect letterforms that collapse under scrutiny. Crowds, water and hands are the motion cases.

SynthID marks every frame invisibly and nobody can strip it. English is the only prompt language Google has evaluated, and Google folded Vertex AI into the Gemini Enterprise Agent Platform on 22 April 2026 — that is where this model is listed for enterprise buyers.

Writing a shot with no picture to lean on

The model has your words and a seed; whatever you leave unsaid becomes its decision. Describe one shot, never a sequence — three beats in eight seconds gets you a bad cut or a blur.

Fix the camera. "Locked-off wide", "slow dolly left", "handheld follow". An unstated camera drifts.

Fix the light. Sodium streetlamp, overcast afternoon, one practical. Lighting separates a deliberate frame from a render.

Fix the sound. You pay for audio whether you planned it or not, so name the ambience and any spoken line — or the model invents a bed that fights your edit.

Then hold the seed and change one clause at a time. Our Gemini Omni prompting guide goes further.

Frequently asked questions

How much does a text-to-video clip cost here?

$0.09 per second of 720p output — $0.72 for the default eight seconds, $0.90 at the ten-second ceiling. A 40% discount on a $0.15 list rate, checked August 28, 2026. Other tiers: $0.0297 at 360p, $0.135 at 1080p, $0.18 at 4K, audio included.

Is this cheaper than calling Google directly?

At all four tiers. Google bills video in output tokens — $17.50 per million, at 1,931 tokens a second for 360p, 5,792 for 720p, 8,688 for 1080p and 17,376 for 4K — which lands at $0.0338 a second, then $0.1014, $0.1520 and $0.3041 as the tiers climb, audio inside each one. Its announcement table rounds those to $0.03, $0.10, $0.15 and $0.30; the unrounded figures are the ones you are billed on. Set against that, we save you roughly twelve percent at 360p, eleven at 720p, eleven again at 1080p, and forty-one percent at 4K. Both ladders take the same 0.33× step down to 360p; the difference is at the top, where Google triples its 720p rate and we double ours.

Is this cheaper than the Omni Flash model it replaces?

No — it costs more. Omni Flash text-to-video runs $0.0825 a second here against $0.09 on this page, so eight seconds is $0.66 there and $0.72 here. Nine percent apart. The older model has no 360p tier and never shipped image-to-video or start-end-frame-to-video on E2X, which is where the difference goes. Calling Google directly is a separate question: its published rate works out to $0.1014 a second at 720p, so both this page and the older one sit under it. The gap is not the product costing more to run: the list rate behind all three is the same $0.15; only the standing discount moves — 40% on this page, 45% and 50% on the older ones.

Should I pick this or Veo 3.1 Fast for a single clip?

Veo 3.1 Fast, almost always — $0.01 a second on the same API, so eight seconds costs $0.08 rather than $0.72. Come back when the clip must reach ten seconds, needs 4K, or has to match other footage.

What is the 360p tier actually for?

Iteration. Four seconds at 360p is $0.1188 against $0.36 at 720p, so the takes you throw away cost a fraction of the one you keep.

How long can a single generation run?

Three to ten seconds, at 24 fps. The duration enum exposes 4, 6, 8 and 10, as strings. Longer pieces come from several calls — what the 1.1 continuity features exist to make survivable.

Does the per-second price include audio?

Yes, on every row of every comparison here. There is no audio toggle at all — dialogue and ambience are generated with the picture, and no provider on that list prices them separately.

Why does another provider quote a different price for this model?

Usually because it is a different model. fal.ai's versionless google/gemini-omni-flash listing prices the deprecated preview weights near $0.125 a second; its v1.1 page reads $0.10, and our table uses v1.1. Check the version and the date first.